Field Trials Using Recycled Glass as Natural Sand

Replacement and Powdered Glass as Cementitious

Materials Replacement in Concrete Pavement

Ion Dumitru1, Tony Song

1, Bob Bornstein

1, Phillip Brooks

2 and Justin Moss

2

1Boral Construction Materials, PO Box 400, Winston Hills, NSW, 2153, Australia

Email: ion.dumitru@boral.com.au, tony.song@boral.com.au, bob.bornstein@boral.com.au 2Road and Maritime Services, Level 6, 101 Miller Street, North Sydney, NSW, 2060,

Australia, email: phillip_Brooks@rms.nsw.gov.au 3Road and Maritime Services, PO Box 3035, Parramatta, NSW, 2124, Australia, email:

Justin_Moss@rms.nsw.gov.au

ABSTRACT

Laboratory investigations, using recycled glass as natural sand replacement and powdered

glass as cementitious materials replacement, have resulted in concrete mixes properties well

within current specification requirement. The major benefits of concrete mixes using crushed

glass as natural sand replacement were lower drying shrinkage and significant lower chloride

diffusion coefficients. The results of this investigation were presented at the second

International Conference on Sustainable Construction Materials and Technologies, in

Ancona, Italy, in 2010.

This paper, documents the results of a field concrete pavement trial, using crushed recycled

glass as natural sand replacement, and powdered glass as cementitious material replacement.

It was concluded that, partial natural sand replacement with crushed recycled glass and,

partial replacement of cementitious materials with powdered glass, can meet the current

concrete pavement specification requirements, with the exception of abrasion resistance

which is lower. A protocol regarding the testing regime, has also been considered, in order to

achieve not only daily concrete production consistency, but also, to address the potential

health and environmental issues.

Keywords. Recycled glass sand, recyced powdered glass, pavement, premix concrete

LABORATORY TRIALS

Natural Sand Replacement

Laboratory trials, carried out with recycled crushed glass as natural sand replacement in

concrete pavement, and, the results of the trials were presented at Ancona, Italy, conference

in 2010 [Dumitru et al. 2010]. The trials were carried out with a clear and green glass with a

size distribution between 3mm – 0.3mm, with a particle density (SSD) of 2.49 t/m3, and with

very low level of chloride and sulphates.

The recycled crushed glass was non-alkali silica reactive. The concrete trial mixes using

30%, 45% and 60% crushed glass, as natural sand replacement, have demonstrated that, the

45% natural sand replacement resulted in higher compressive strength at 28 days, when

compared with the control mix using 100% natural sand. Results are presented in Figure 1.

Drying shrinkage, AVPV (apparent volume of permeable voids – AS 1012.21) were also

lower when compared with the control.

When assessed for durability, in terms of chloride diffusion coefficient as per NT443, the

concrete mix using 45% natural sand replacement has performed better than other mixes as

presented in Figure 2.

Figure 1. Compressive strength using glass sand replacement

1.20

1.30

1.40

1.50

1.60

1.70

1.80

Control (no glass) 30% sand

replacement

45% sand

replacement

60% sand

replacement

Chloride diffusion (x10-11)

Figure 2. Chloride Diffusion Coefficient of laboratory concrete trials

70%

75%

80%

85%

90%

95%

100%

105%

110%

1 7 28 56 90

Concrete age (day)

Percentage (%)

Control (no glass) 30% sand replacement45% sand replacement 60% sand replacement

Cementitious Materials Replacement

Powdered glass with a density of 2.48 t/m3, a Fineness Index of 335m

2/kg, and a mean

diameter of 54.1 microns, has been used in the concrete pavement trials mixes (see Figure 3),

using 7.5% - 15% replacement.

The finding of the trial mixes carried out, have concluded that:

• Setting time is up to 1-2 hours longer

• Air content is higher than control

• Compressive strength is lower at 28 days than control

• Abrasion resistance is lower

It is important to note that, the mixes using powdered glass as cementitious materials

replacement, have met the design criteria for concrete pavement compressive strength and,

for the field trial, a 15% cementitious materials replacement has been considered.

0

1

2

3

4

5

6

0.01 0.1 1 10 100 1000 10000

Particle s ize (um)

Volume (%)

Figure 3. Particle Size Distribution of Glass Powdered

FIELD TRIALS

The field trials using recycled crushed glass, and powdered glass in concrete pavement, have

been carried out at Boral’s Dunmore Quarry, South of Sydney, using control mix cap, and a

control mix between the glass sand concrete and the powdered glass concrete (Figure 4).

The mix design for the trials, has used 45% replacement of natural sand with recycled

crushed glass, and approximately 15% powdered glass replacement of cementitious

materials. Although the laboratory trials have demonstrated that 30% recycled glass has

performed better, the 45% replacement of natural sand with recycled glass has been chosen

in order to assess the field performance of a concrete pavement using large quantities of

recycled crushed glass.

Details of the mix designs are presented in Table 1, the mix using cement and fly ash.

The pavement design was a typical concrete road pavement used in main roads in New South

Wales, with a lean mix subbase, two coat seal and, proper anchor trench steel, as shown in

Figure 5 .

Figure 4. Field trials plan

Table 1. Concrete Mix Design used for Field Trials

Mix Control Glass Sand 45% Glass Powder 15%

CEMENTITIOUS kg/m3 395 397 337

Glass Powder kg/m3 - - 60

COARSE Sand (DSS) kg/m3 664 367 664

Glass Sand kg/m3 - 292

10mm Dunmore kg/m3 325 327 325

20mm Dunmore kg/m3 749 755 749

WATER (incl. Admix.) kg/m3 185 175 185

AEA-940 ml/m3 300 200 300

POZZ-300RI ml/100kg 200 200 200

Figure 5. Anchor trench steel / formwork for field trials

Recycled Crushed Glass as natural Sand Replacement

The 35MPa mix using 45% recycled crushed glass, has a slightly lower slump than the

control, with the air content also slightly lower than control. The compressive strength of the

mixing, using recycled glass as natural sand replacement, has been measured in cylinders

taken from concrete delivered in five trucks, with all results higher than the design 35MPa.

At 91 days, the compressive strength was almost 55MPa, well within the specification

requirements.

The results for compressive strength at 28 and 91 days are presented in Figure 6.

20.0

25.0

30.0

35.0

40.0

45.0

50.0

55.0

60.0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Truck Load No

28d Compressive strength (MPa)

Glass sand concrete Control concrete Powder Glass concrete

Minimum 35MPa@28d

Figure 6 Concrete compressive strength in cylinders from field trials at 28d

(colour) and 91d (diagonal line)

The flexural strength, of the concrete using crushed glass, was lower than 4.5MPa

requirements at 28 days, but it was as high as 6MPa at 91 days.

As expected, the concrete shrinkage at 21 days was very low at around 420 microstrain and,

was less than 600 microstrain at 56 days. The results are lower than the control, as presented

in Figure 7.

Interesting results were obtained for compressive strength, on cores drilled from site at age

of 28 days.

The data, as presented in Figure 8, show that all results of concrete cores are above the

minimum requirements of 35MPa, with some compressive strength as high as 45MPa, the

average of 6 cores being almost 40MPa, slightly higher than the control mix.

220

320

420

520

620

720

820

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Truck Load No

Drying shrinkage (micronstrain)

Glass sand concrete Control concrete Powder Glass concrete

Figure 7 Concrete drying shrinkage of concrete from field trials at 21d (colour)

and 56d (diagonal line)

35.1

41.8

36.2

42.0

37.8

28.7 28.2

31.8

39.9

34.3

31.5

45.4

41.5

39.6

38.440.4

33.2

35.7

25

30

35

40

45

50

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Concrete Core No

Compressive Strength (MPa)

Glass sand

Control

Powder Glass

compaction issue

(?) as cores have

been taken from

edge areas

Figure 8. Compressive strength @28d of cores from field concrete

Powdered Glass as Cementitious Material Replacement

The results are a true reflection of the laboratory trials. About 15% cementitious material

was replaced with powdered glass, with a fineness index of 335m2/kg, resulting in lower

compressive strength in cylinders and cores, presented in Figures 6 and 8, respectively.

However, at 91 days, the compressive strength was higher than 45MPa in concrete cylinders.

Some lower results in cores, as presented in Figure 8, are due to compaction issue, as the

cores were taken from the concrete edge areas.

The concrete shrinkage at 21 and 56 days are higher than control, as presented in Figure 7.

ASSESSMENT OF FIELD TRIAL AREA AFTER ONE YEAR IN SERVICE

Visual inspection, carried out after one year in service, has concluded that the pavement is in

a good condition, with some aggregate loss of about 20-30% for both areas, using glass as

natural sand replacement and powdered glass as cementitious materials replacement.

As expected, the abrasion resistance is lower than the control for both trials, being obvious

that the control performed better, as presented in Figures 9 and 10.

Figure 9. Side-by-Side Comparison of Surface Wear in IWP – Control vs Glass

Sand (1x & 3x)

Figure 10. Side-by-Side Comparison of Surface Wear in IWP – Control vs

Powdered Glass (1x & 3x)

Concrete cylinders tested at 365 days from crushed glass area, showed compressive strengths

of 62.9MPa and, 50.8MPa for powdered glass area.

With regard to the visual inspection carried out, it is important to note that:

• Wheels and windborne grit from the quarry are prematurely accelerating abrasion.

• With coarse aggregate now exposed, the rate of abrasion is slow.

• Compressive strength does not correlate well with wear resistance.

The monitoring will continue on annual basis.

Based on the good results obtained in the field trials, glass sand is currently being used in

production at 20-25% of the total sand content. After approximately 5 months, the results,

before and after using glass, indicate no statistical differences at 28 days between concrete ,

with and without glass for 20, 25, 32 and 40MPa grade.

A protocol, regarding the testing regime for glass sand and powdered glass/natural sand

blends has been introduced for concrete production. The protocol also covers health and

technical aspects including dust, grading, chloride, lead, sulphur trioxide, sugar, setting time,

etc.

CONCLUSIONS

• Recycled sand glass can be used to partially replace natural sand in concrete,

producing concrete with at least equivalent fresh and hardened properties;

• The drying shrinkage, chloride diffusion coefficient and AVPV (apparent volume of

permeable voids), decreases with the increase in replacement percentage of natural

sand;

• Abrasion resistance of concrete with recycled glass is lower, and this is a major

hurdle to be addressed in concrete pavement;

• Powdered glass with a Fineness Index of 335m2/kg can be used to partially replace

up to 15% of cementitious materials in concrete pavements;

• The compressive strength of the concrete pavement has achieved the minimum

35MPa as per specification requirements. Flexural strength has achieved the required

4.5MPa at 91 days;

• Drying shrinkage at 56 days is slightly higher than the control;

• Abrasion resistance is lower and setting time is longer when compared to control

mix.

REFERENCES

Australian Standards AS1012.21, “Determination of Water Absorption and Apparent Volume

of Permeable Voids in Hardened Concrete”, Australia, 9 pages

Dumitru, I., Song, T., Caprar, V., Brooks, P., and Moss, J., (2010), “Incorporation of

recycled glass for durable concrete”, Second International Conference on

Sustainable Construction Materials and Technologies, Universita Polytechnica delle

Marche, Ancona, Italy, 9 pages

Nordtest Method NT BUILD 443 (1995), “Concrete, hardened: Accelerated Chloride

Penetration”, ISSN 0283-7153, 5 pages